Data grading method and device, equipment, storage medium and computer program

Abstract

The invention relates to a data grading method and device, equipment, a storage medium and a computer program. The method comprises the steps of obtaining to-be-graded target data; aiming at the target data, endowing each layer of index in a preset hierarchical data grading index system with a weight; determining a data classification set; scoring each dimension index of a bottom layer in the data grading index system based on the fuzzy membership degree to form a corresponding membership degree vector of each dimension index of the bottom layer, and forming an evaluation matrix of each dimension index of the bottom layer based on the membership degree vector of each dimension index of the bottom layer; performing fuzzy transformation to obtain a fuzzy evaluation set; and determining the level corresponding to the maximum element in the fuzzy evaluation set as a grading result of the target data. Through multi-level and multi-dimensional index refinement, scientific and reasonable grading processing of the data can be realized, and rationality application and safety protection of the data are facilitated.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular to a data classification method, device, equipment, storage medium and computer program. Background Art

[0002] In view of the data classification and grading protection system, it is required to establish data classification and grading protection from different levels and dimensions, and the requirements for data security protection are increasing year by year. Classification and grading are important means of data security protection, strengthening the protection of important data and implementing stricter management of core data.

[0003] In the practice of data security protection based on data classification and grading, challenges are still faced. In the process of implementing data grading, the relevant standards do not provide specific implementation methods and relevant organizational evaluation mechanisms for data grading. There are problems such as non-standardization, unprofessionalism, and non-representativeness in the decision-making process of grading. In addition, the generalization and lack of specificity of the grading evaluation process mechanism leads to poor scientificity, rationality, consistency, and objectivity of the grading results, which in turn will lead to risks in the rational application of data. Summary of the invention

[0004] The present disclosure provides a data classification method, apparatus, device, storage medium and computer program to achieve scientific and reasonable classification processing of data, which is conducive to the rational application and security protection of data.

[0005] In a first aspect, the present disclosure provides a data classification method, comprising:

[0006] Obtain target data to be classified;

[0007] For the target data, weights are assigned to indicators at each level in a preset hierarchical data classification indicator system, wherein the data classification indicator system includes at least first-level indicators, second-level indicators and third-level indicators at different levels, wherein the first-level indicators are data classification indicators, the second-level indicators include qualitative data importance indicators, quantitative data importance indicators and data security risk impact indicators, and the third-level indicators include lower-level indicators of the second-level indicators;

[0008] determining a data hierarchical set, the data hierarchical set comprising a plurality of levels;

[0009] Scoring the underlying dimensional indicators in the data classification indicator system based on fuzzy membership to form corresponding underlying dimensional indicator membership vectors, and forming an underlying dimensional indicator evaluation matrix based on the underlying dimensional indicator membership vectors;

[0010] The weight vector formed by the weights of the indicators of each dimension in the bottom layer is fuzzily transformed with the evaluation matrix of the indicators of each dimension in the corresponding bottom layer by using the fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the second bottom layer; in the second bottom layer, the evaluation matrix of the indicators of each dimension in each corresponding layer is formed by the membership vector of the indicators of each dimension in each same layer, and the weight vector formed by the weights of the indicators of each dimension in each corresponding layer is fuzzily transformed with the evaluation matrix of the indicators of each dimension in each corresponding layer by using the fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, and the evaluation matrix of the indicators of each dimension in the previous layer of each corresponding layer is formed based on the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, iterate upward along the data classification indicator system to the second layer, and obtain the fuzzy judgment set by fuzzy transformation of the weight vector formed by the weights of the indicators of each dimension in the second layer and the evaluation matrix of the indicators of each dimension in the second layer;

[0011] The level corresponding to the maximum element in the fuzzy evaluation set is determined as the classification result of the target data.

[0012] In some embodiments, for the target data, weights are assigned to indicators at each level in a preset hierarchical data classification indicator system, including:

[0013] Through iterative voting of important indicators by expert scoring and statistical election density, weights are assigned to indicators at each level in the data grading indicator system.

[0014] In some embodiments, based on the expert scoring, iterative voting and election density statistics, weights are assigned to each level of the preset hierarchical data classification index system, including:

[0015] Determine the index sequence number set SN of the current layer and determine the index set F of the current layer;

[0016] Determine the expert group and its number. Experts who are required to vote in each round of voting may not abstain and must cast one vote;

[0017] Construct a valid voting indicator set E and initialize related parameters to enter the next round of iteration;

[0018] In the rth round, the set of all experts who voted for the index with the maximum voting density in the r-1th round is selected to vote on the most important index in E. In the r-1th round, the votes of experts other than the set of experts who voted for the index with the maximum voting density in the rth round are still valid in the rth round, and these experts will not vote again in the rth round;

[0019] Count and store the votes for each indicator in round r of E;

[0020] Based on the number of elements in E in the rth round and r-1th round, the voting correction coefficient of each indicator in E in the rth round is calculated and stored;

[0021] Based on the maximum voting density of each indicator in E in the r-1th round, the voting correction coefficient of each indicator in E in the rth round, the number of votes obtained by the corresponding indicator in F in E in the rth round iteration, and the number of experts, the voting density of each indicator in E in the rth round is calculated and stored;

[0022] Get the set of serial numbers of the index with the largest voting density in E in the rth round in F, correspond the index with the largest voting density in E in the rth round to F, add the index corresponding to the maximum voting density and the elements in the tuple set formed by the voting density to the election density tuple set S;

[0023] Store the maximum voting density of each indicator in the rth round E, and delete the indicator and its serial number corresponding to the maximum voting density in the rth round;

[0024] Determine whether the number of elements in E is 1:

[0025] If it is 1, calculate and store the voting density of the last indicator in E, and form a tuple set of the maximum voting density indicator and density, and add the only indicator in the tuple set to the election density tuple set S;

[0026] Otherwise, determine whether the element in E is 0:

[0027] If it is 0, the voting density vp in the tuples in the election density tuple set S is normalized to obtain the election density of each indicator, and the election density of each indicator is determined as the weight of each indicator;

[0028] Otherwise, proceed to the next iteration.

[0029] In some embodiments, the calculation formula of the voting correction coefficient is as follows:

[0030]

[0031] In the formula, c r Indicates the correction coefficient for the rth round of voting, elnum r Indicates the number of elements in the rth round E, elnum (r-1) Represents the number of elements in E in the r-1th round;

[0032] The calculation formula for the voting density of each indicator in the rth round E is as follows:

[0033]

[0034] In the formula, vp rj represents the density of the jth index in F in E that is voted in the rth iteration, vpmax (r-1) represents the maximum value of the voting density of each indicator in the r-1th round E, and n represents the number of experts in the expert group;

[0035] The election density of each indicator is determined as the weight of each indicator, using the following calculation formula:

[0036]

[0037] In the formula, sp j represents the election density of the j-th indicator in F, S[j][2] represents the second element value of the j-th tuple in S, and S[x][2] represents the second element value of the x-th tuple in S.

[0038] In some embodiments, the underlying dimensional indicators in the data classification indicator system are scored based on the fuzzy membership to form corresponding underlying dimensional indicator membership vectors, including:

[0039] Based on the importance of the target data in the underlying dimensional indicators under the qualitative data importance index, vote on the importance of the corresponding underlying dimensional indicators, count and quantify the voting results, and form the corresponding underlying dimensional indicator membership vector;

[0040] Based on the impact degree of the target data on the underlying dimensional indicators under the data security risk impact degree indicator, vote on the risk impact degree of the corresponding underlying dimensional indicators, count and quantify the voting results, and form the corresponding underlying dimensional indicator membership vector;

[0041] Get the values ​​of the underlying dimensional indicators under the quantitative data importance index, use the preset membership function corresponding to each level to calculate the output value of the corresponding level and normalize the output value to form the corresponding membership vector of the underlying dimensional indicators.

[0042] In some embodiments, the preset membership function includes one or more of an S-type membership function, a double S-type membership function, a Z-type membership function, a trapezoidal membership function, a triangular membership function, a Gaussian membership function, a Gaussian combination membership function, and a generalized bell-shaped membership function.

[0043] In a second aspect, the present disclosure provides a data classification device, comprising:

[0044] An acquisition module, used to acquire target data to be classified;

[0045] A weighting module, for assigning weights to indicators at each level in a preset hierarchical data classification indicator system for the target data, wherein the data classification indicator system includes at least primary indicators, secondary indicators and tertiary indicators at different levels, wherein the primary indicators are data classification indicators, the secondary indicators include qualitative data importance indicators, quantitative data importance indicators and data security risk impact indicators, and the tertiary indicators include lower-level indicators of the secondary indicators;

[0046] A first determining module, configured to determine a data hierarchical set, wherein the data hierarchical set includes a plurality of levels;

[0047] A scoring module is used to score the underlying dimensional indicators in the data classification indicator system based on fuzzy membership, form corresponding underlying dimensional indicator membership vectors, and form an underlying dimensional indicator evaluation matrix based on the underlying dimensional indicator membership vectors;

[0048] A fuzzy transformation module is used to perform fuzzy transformation on the weight vector formed by the weights of the indicators of each dimension in the bottom layer and the evaluation matrix of the indicators of each dimension in the corresponding bottom layer through a fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the second bottom layer; in the second bottom layer, the evaluation matrix of the indicators of each dimension in each corresponding layer is formed by the membership vector of the indicators of each dimension in each same layer, and the weight vector formed by the weights of the indicators of each dimension in each corresponding layer and the evaluation matrix of the indicators of each dimension in each corresponding layer are fuzzy transformed through a fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, and the evaluation matrix of the indicators of each dimension in the previous layer of each corresponding layer is formed based on the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, iterate upward along the data classification indicator system to the second layer, and obtain the fuzzy judgment set through the fuzzy transformation of the weight vector formed by the weights of the indicators of each dimension in the second layer and the evaluation matrix of the indicators of each dimension in the second layer;

[0049] The second determination module is used to determine the level corresponding to the maximum element in the fuzzy evaluation set as the classification result of the target data.

[0050] In a third aspect, the present disclosure provides a computer device, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the first aspect.

[0051] In a fourth aspect, the present disclosure provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described in the first aspect when executed by a processor.

[0052] In a fifth aspect, the present disclosure provides a computer program product, comprising a computer program / instructions, which implements the steps of the method described in the first aspect when the computer program is executed by a processor.

[0053] The present invention provides a data classification method, device, equipment, storage medium and computer program, which quantify the classification basis in terms of qualitative and quantitative aspects of data importance and risk impact of data through multi-level and multi-dimensional indicator refinement, avoids the inconsistency problem of subjective weighting methods, simplifies and integrates group decision-making mechanisms, and generates weight results through deterministic iterative optimization. This solution has low time and space complexity and is easy to accept, understand and implement. It can realize scientific and reasonable classification processing of data, which is beneficial to the reasonable application and security protection of data. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The present disclosure will be described in more detail below based on embodiments and with reference to the accompanying drawings:

[0055] Figure 1 A flowchart of a data classification method provided in an embodiment of the present disclosure.

[0056] Figure 2 A schematic diagram of a data grading indicator system provided for an embodiment of the present disclosure.

[0057] Figure 3 A schematic diagram of scoring and grading qualitative data importance indicators provided in an embodiment of the present disclosure.

[0058] Figure 4 A schematic diagram of scoring and grading the impact degree indicators of data security risks provided in an embodiment of the present disclosure.

[0059] Figure 5 A schematic diagram of scoring and grading the impact degree indicators of data security risks provided in an embodiment of the present disclosure.

[0060] Figure 6 A schematic diagram of the evaluation matrix and weight distribution of the indicator system tree provided in the embodiment of the present disclosure.

[0061] Figure 7 A schematic diagram of an application example provided for an embodiment of the present disclosure.

[0062] In the drawings, the same reference numerals are used for the same components, and the drawings are not drawn to scale. DETAILED DESCRIPTION

[0063] In order to enable those skilled in the art to better understand the technical solution of the present disclosure, and to fully understand and implement how the present disclosure applies technical means to solve technical problems and achieve the corresponding technical effects, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only embodiments of a part of the present disclosure, not all of the embodiments. The embodiments of the present disclosure and the various features in the embodiments can be combined with each other without conflict, and the technical solutions formed are all within the scope of protection of the present disclosure. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present disclosure.

[0064] Embodiment 1

[0065] Figure 1 The following is a flow chart of a data classification method provided by an embodiment of the present disclosure. Figure 1 As shown, this embodiment provides a data classification method, including:

[0066] Step S11, obtaining target data to be classified.

[0067] In practical applications, the target data can be data from any industry field.

[0068] Step S12: for the target data, weights are assigned to the indicators at each level in the preset hierarchical data grading indicator system. The data grading indicator system includes at least first-level indicators, second-level indicators and third-level indicators at different levels, wherein the first-level indicators include data grading indicators, the second-level indicators include qualitative data importance indicators, quantitative data importance indicators and data security risk impact indicators, and the third-level indicators include the lower-level indicators of the second-level indicators.

[0069] The data classification index system in this embodiment is divided into categories based on the importance of data in economic and social development, and the degree of harm caused to national security, economic operations, social order, public interests, organizational rights and interests, and personal rights and interests when the data is subject to security risks such as leakage, tampering, damage, or illegal acquisition, illegal use, and illegal sharing.

[0070] like Figure 2As shown in the figure, the data classification indicator system includes indicators of different levels. Indicators of different levels constitute a hierarchical data classification indicator system. The first-level indicators are data classification indicators, and the lower levels are divided into qualitative data importance indicators, quantitative data importance indicators and data security risk impact indicators. At the same time, under these three types of indicators, the impact indicators of data classification are refined in multiple levels and dimensions, and data classification is carried out after identifying the above indicators.

[0071] Figure 2 In the data, the qualitative data importance index is divided into three types of indicators: business field, application entity and region. Among them, the business field is further divided into industry, business and content, the application entity is further divided into three levels of indicators such as people, groups, groups and systems, and the region is further divided into four levels of indicators such as places, geographical areas, administrative areas and socio-economic fields. The quantitative data importance index is divided into three levels of indicators such as accuracy, scale and coverage. The accuracy is further divided into four levels of indicators such as time accuracy and space accuracy. The scale is further divided into four levels of indicators such as storage volume, transaction volume and transmission volume. The coverage is further divided into four levels of indicators such as industry field, audience, geographical area and time. The impact of data security risks is further divided into three levels of indicators such as leakage, tampering, damage, illegal acquisition, illegal use and illegal sharing. The hierarchical data classification index system constructed based on these multi-level and multi-dimensional refined indicators can be applied to the data classification processing in this embodiment, so as to obtain scientific and accurate classification identification results, providing a basis for reasonable and effective data application and protection.

[0072] In some specific implementations, for target data, weights are assigned to indicators at each level in a preset hierarchical data classification indicator system, including:

[0073] Step S12a: weight the indicators at each level in the data classification indicator system by iteratively voting on important indicators through expert scoring and statistical election density.

[0074] Further, step S12a may include:

[0075] Step S12a-1, determine the indicator sequence number set SN = {i|i = 1, ...m} of the current layer, determine the indicator set F = {fi|i∈SN} of the current layer, m represents the number of indicators in F, fi represents the i-th indicator in F.

[0076] Step S12a-2: Determine the expert group and the number of people n. In each round of voting, the experts who need to vote cannot abstain and must cast one vote.

[0077] Step S12a-3, construct a valid voting indicator set E, initialize E to F, E=F, initialize the indicator serial number set ESN in E to SN; initialize the election density tuple set S={}, the number of iteration rounds of expert voting r=0; store the number of elements in E in the 0th round elnum0=m, the number of elements in E in the 1st round elnum1=m; the maximum voting density vpmax0=1 in E in the 0th round, and the set ex0 of all experts who voted for the corresponding indicator with the maximum voting density in the 0th round.

[0078] Execute r=r+1 and enter the next round of iteration.

[0079] E={erj|j∈ESN}, erj represents the jth index in F in E in the rth round.

[0080] Step S12a-4: In the rth round, select the set of all experts ex corresponding to the index with the maximum voting density in the r-1th round (r-1) Vote for the most important indicator in E. In the r-1th round, except for the expert set ex corresponding to the indicator with the maximum voting density in the rth round, (r-1) The votes of experts other than the above experts are still valid in round r, and these experts will not vote again in round r.

[0081] Step S12a-5: Count and store the number of votes for each indicator in the rth round E = {v rj |j∈ESN},v rj It represents the number of votes obtained by the jth indicator in F in E in the rth round.

[0082] Step S12a-6: Based on the number of elements in the rth round and the r-1th round E, the voting correction coefficient of each indicator in the rth round E is calculated and stored.

[0083] In some specific implementations, the calculation formula for the voting correction coefficient is as follows:

[0084]

[0085] In the formula, c r Indicates the correction coefficient for the rth round of voting, elnum r Indicates the number of elements in the rth round E, elnum (r-1) Represents the number of elements in E in the r-1th round.

[0086] Step S12a-7, calculate and store the voting density of each indicator in E in the r-1th round based on the maximum voting density of each indicator in E in the r-1th round, the voting correction coefficient of each indicator in E in the rth round, the number of votes obtained by the corresponding indicator in F in E in the rth round iteration, and the number of experts in the expert group.

[0087] The calculation formula for the voting density of each indicator in the rth round E is as follows:

[0088]

[0089] In the formula, vp rj represents the density of the jth index in F in E that is voted in the rth iteration, vpmax (r-1) It represents the maximum voting density of each indicator in the r-1th round E, and n represents the number of experts in the expert group.

[0090] Step S12a-8, obtaining the set of serial numbers of the index with the largest voting density in round E in F, Among them, mx r It represents the set of numbers in F of the index with the highest voting density in E in the rth round. The index with the highest voting density in E in the rth round corresponds to ec in F. r ={e rg |g∈mx r}, add the index corresponding to the maximum voting density and the elements in the tuple set formed by the voting density to the election density tuple set S, and the index corresponding to the maximum voting density and the tuple set formed by the voting density are recorded as et r ={(e rg ,vp rg )|g∈mx r}.

[0091] Step S12a-9: Store the maximum voting density vpmax of each indicator in the rth round E r , Among them, mx r [0] indicates mx r The first element of the set, so vp r(mxr[0]) represents the maximum voting density in the rth round E;

[0092] Delete the index and its serial number corresponding to the maximum value of the voting density in round r: Delete ec from E r , E=E-ec r , remove mx from ESN r ,ESN=ESN-mx r ;

[0093] elnum (r+1) =||E||, the number of elements in E is saved to elnum (r+1) .

[0094] Step S12a-10: Determine whether the number of elements in E is 1:

[0095] If it is 1, execute step S12a-11;

[0096] Step S12a-11, calculate and store the voting density of the last indicator in E, and form a tuple set et of the maximum voting density indicator and density. (r+1) , et (r+1) =(e rlst ,vp rlst ), where lst represents the index number of the only index in E in F, et (r+1) In the rth round of calculation, the tuple set et (r+1) The unique index in is added to the election density tuple set S;

[0097] Otherwise, execute step S12a-12;

[0098] Step S12a-12, determine whether the element in E is 0:

[0099] If it is 0, execute step S12a-13; otherwise, execute r=r+1, enter the next round of iteration, and execute step S12a-4;

[0100] Step S12a-13: normalize the voting density vp in the tuples in the election density tuple set S to obtain the election density of each indicator, and determine the election density of each indicator as the weight of each indicator.

[0101] In some specific implementations, the election density of each indicator is determined as the weight of each indicator, using the following calculation formula:

[0102]

[0103] In the formula, sp j represents the election density of the j-th indicator in F, S[j][2] represents the second element value of the j-th tuple in S, and S[x][2] represents the second element value of the x-th tuple in S.

[0104] It should be understood that the above only describes the weighting process of the same level of indicators. In this embodiment, each level of indicators in the data classification indicator system can be weighted by the above-mentioned expert scoring, iterative voting, and statistical voting density. No further details will be given here.

[0105] Step S13: determine a data hierarchical set, where the data hierarchical set includes multiple levels.

[0106] In this embodiment, the level is set to 5 levels, L = {level A, level B, level C, level D, level E}

[0107] The meaning of data importance indicators (including qualitative and quantitative indicators) is as follows:

[0108] Indicator level meaning Level A Very important Class B important C-Class More important D-Class generally E-Class unimportant

[0109] For the risk impact index, its meaning is:

[0110]

[0111]

[0112] Step S14: score the underlying dimensional indicators in the data classification indicator system based on the fuzzy membership to form corresponding underlying dimensional indicator membership vectors, and form an evaluation matrix for each dimensional indicator based on the underlying dimensional indicator membership vectors.

[0113] In one example, the membership vector elements are as follows:

[0114] Level A Class B C-Class D-Class E-Class 0.1 0.3 0.4 0.2 0

[0115] In some specific implementations, the underlying dimensional indicators in the data classification indicator system are scored based on the fuzzy membership to form corresponding underlying dimensional indicator membership vectors, including:

[0116] Step S14a: Based on the importance of the target data in the underlying dimensional indicators under the qualitative data importance index, vote on the importance of the corresponding underlying dimensional indicators, count and quantify the voting results, and form the corresponding underlying dimensional indicator membership vector.

[0117] For qualitative data importance index items, we use expert voting to count the votes at each level and normalize the total. Figure 3 As shown in the figure, the evaluators vote on the importance of the indicator items at five levels: A, B, C, D, and E, based on the importance of the target data in the corresponding indicator items. The voting results are counted and quantified to form a membership vector. For example, the membership vector VI corresponding to the indicator item "Industry" is calculated as follows:

[0118]

[0119] Where NI A ,NI B ,NI C NI D NI E They represent the importance of the current target data in the industry, the number of votes at level A, level B, level C, level D, and level E, and each evaluator has one vote. Na is the number of evaluators. Evaluators are not allowed to abstain from voting. The sum of the elements of the membership vector is 1, which means that the total is normalized. The formation of other vectors is similar.

[0120] Step S14b: Based on the impact of the target data on the underlying dimensional indicators under the data security risk impact index, vote on the risk impact of the corresponding underlying dimensional indicators, count and quantify the voting results, and form the corresponding underlying dimensional indicator membership vector.

[0121] like Figure 4 As shown in the figure, for the impact index items of data security risks, the assessors vote on the risk impact of the index items at five levels: A, B, C, D, and E, based on the impact of the target data on the corresponding index items. The voting results are counted and quantified to form a membership vector. For example, the membership vector VL corresponding to the index item "leakage" is calculated as follows:

[0122]

[0123] In the formula, NL A NL B NL C NL D NL E They represent the degree of adverse impact caused by the leakage of the current target data, the number of votes at level A, level B, level C, level D, and level E, and one vote for each evaluator. Evaluators are not allowed to abstain from voting. The sum of the elements of the membership vector is 1, that is, the sum is normalized. The formation of other vectors is similar.

[0124] Step S14c, obtain the numerical values ​​of the underlying dimensional indicators under the quantitative data importance index, use the preset membership function corresponding to each level to calculate the output value of the corresponding level and normalize the output value to form the corresponding membership vector of the underlying dimensional indicators.

[0125] For quantitative data importance indicators, the corresponding small-scale membership function, large-scale membership function or intermediate-scale membership function can be selected and configured according to the value range of the indicator data and the physical meaning of the indicator value. Figure 5As shown in the figure, the underlying indicator items under the impact degree index of data security risk belong to objective evaluation items, and the indicator levels can be divided into 5 levels, namely A, B, C, D, and E. Scoring and grading are achieved by presetting the membership functions of each indicator item at A, B, C, D, and E, and calculating the level membership corresponding to the indicator data. The preset membership function can select a small-sized membership function, a large-sized membership function, or an intermediate-sized membership function according to the nature of the indicator item. Specifically, one or more of the S-type membership function, double S-type membership function, Z-type membership function, trapezoidal membership function, triangle membership function, Gaussian membership function, Gaussian combination membership function, and generalized bell-shaped membership function can be selected. For one indicator, 5 levels need to set 5 membership functions, and membership functions of uniform shape can be selected, or membership functions of different shapes can be selected according to specific needs. For example, for one indicator, the indicator value is input into the corresponding five levels of membership function, and five values ​​are obtained, which correspond to five values ​​of different levels of the membership vector. Then, the values ​​of the vector elements must be normalized by summing them.

[0126] In this embodiment, for a quantitative data importance index item, a membership function is preset at five levels: A, B, C, D, and E. Based on the specific index value obtained as input, the output value at the corresponding level is calculated and normalized to form the corresponding underlying dimensional index membership vector. For example, the membership vector VDS corresponding to the "storage capacity" index is calculated as follows:

[0127] SDS=mfDS A (x)+mfDS B (x)+mfDS C (x)+mfDS D (x)+mfDS E (x)

[0128]

[0129] Where x is the value of the "storage capacity" indicator, mfDS A 、mfDS B 、mfDS C 、mfDS D 、mfDS E Respectively represent the preset membership functions at five levels: A, B, C, D, and E. mfDS A (x), mfDS B (x), mfDS C (x), mfDS D (x), mfDS E(x) represents the function value of each membership function with x as input, and SDS represents the sum of the function values ​​of the preset membership functions at five levels: A, B, C, D, and E. The sum of the elements of the membership vector is 1, that is, the total is normalized. The formation of other vectors is similar.

[0130] After the above process, this embodiment can obtain Figure 6 The evaluation matrix and weight distribution diagram of the indicator system tree shown in Figure 2 are shown in Figure 2. After the membership vector is formed, the evaluation matrix is ​​formed based on the membership vector, as shown in Figure 2. Figure 6 As shown, R1 T is the transpose of the evaluation matrix R1, and so on. The row vectors in R1 are the membership vectors of the underlying dimensional indicators in the business field. For example, the row vectors in R1 correspond to the membership vectors of the three indicators of industry, business and content, and so on. The elements in the weight vector W1 are the weights of the underlying dimensional indicators in the business field, and so on.

[0131] Step S15, fuzzy transformation is performed through a fuzzy synthesis operator to obtain a fuzzy judgment set: a weight vector formed by the weights of the indicators of each dimension in the bottom layer is fuzzily transformed with the evaluation matrix of the indicators of each dimension in the corresponding bottom layer through a fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the second bottom layer; in the second bottom layer, the evaluation matrix of the indicators of each dimension in each corresponding layer is formed with the membership vector of the indicators of each dimension in each same layer, and a weight vector formed by the weights of the indicators of each dimension in each corresponding layer is fuzzily transformed with the evaluation matrix of the indicators of each dimension in each corresponding layer through a fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, and an evaluation matrix of the indicators of each dimension in the previous layer of each corresponding layer is formed based on the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, and the data grading indicator system is iterated upward to the second layer, and the fuzzy judgment set S is obtained through the fuzzy transformation of the weight vector formed by the weights of the indicators of each dimension in the second layer and the evaluation matrix of the indicators of each dimension in the second layer.

[0132] In some specific embodiments, Figure 6 Taking the data classification indicator system shown in the figure as an example, the fuzzy transformation process of the weight vector formed by the weight of the indicators of each layer and each dimension and the evaluation matrix is ​​as follows:

[0133]

[0134] Among them, S1 represents the membership vector of the qualitative data importance index, W8 represents the weight vector formed by the weights of each dimensional index (business field, application entity, region) under the qualitative data importance index, R1 represents the evaluation matrix formed by the membership vector of the underlying dimensional indicators (industry, business, content) under the business field, W1 represents the weight vector formed by the weights of the underlying dimensional indicators under the business field, R2 represents the evaluation matrix formed by the membership vector of the underlying dimensional indicators (people, groups, groups, systems) under the application entity, W2 represents the weight vector formed by the weights of the underlying dimensional indicators under the application entity, R3 represents the evaluation matrix formed by the membership vector of the underlying dimensional indicators (places, geographical areas, administrative areas, socio-economic fields) under the region, W3 represents the weight vector formed by the weights of the underlying dimensional indicators under the region, is the fuzzy synthesis operator.

[0135]

[0136] Among them, S2 represents the membership vector of the quantitative data importance index, W9 represents the weight vector formed by the weights of each dimensional index (precision, scale, coverage) under the quantitative data importance index, R4 represents the evaluation matrix formed by the membership vectors of the underlying dimensional indicators (time accuracy, spatial accuracy) under accuracy, W4 represents the weight matrix formed by the weights of the underlying dimensional indicators under accuracy, R5 represents the evaluation matrix formed by the membership vectors of the underlying dimensional indicators (storage volume, transaction volume, transmission volume) under scale, W5 represents the weight vector formed by the weights of the underlying dimensional indicators under scale, R6 represents the evaluation matrix formed by the membership vectors of the underlying dimensional indicators (industry field, audience, geographical area, time) under coverage, and W6 represents the weight vector formed by the weights of the underlying dimensional indicators under coverage.

[0137]

[0138] Among them, S3 represents the impact index of data security risk ( Figure 6 W7 represents the weight vector formed by the weights of the underlying dimensional indicators (leakage, tampering, damage, illegal acquisition, illegal use, illegal sharing) under the data security risk impact index, and R7 represents the evaluation matrix formed by the membership vectors of the underlying dimensional indicators under the data security risk impact index.

[0139]

[0140] Among them, S represents the fuzzy judgment set of data classification indicators, and W10 represents the weight vector formed by the weights of each dimensional indicator under the data classification indicator (quantitative data importance indicator, qualitative data importance indicator and data security risk impact indicator).

[0141] In this embodiment, by and Obtain the membership vector of the sub-bottom-level dimensional indicators (business field, application entity, region, etc.) of the impact degree of data security risk. Since there is only one layer of multi-dimensional indicators (leakage, tampering, damage, illegal acquisition, illegal use, illegal sharing) below the impact degree of data security risk, the impact degree of data security risk is the second layer. Calculate the membership vector of the impact degree of data security risk Furthermore, for the branches of qualitative data importance index and quantitative data importance index, the weight vector W8 formed by the weights of the third layer of business field, application entity and region, and the weight vector W9 formed by the weights of accuracy, scale and coverage are fuzzily transformed with the corresponding evaluation matrix through the fuzzy synthesis operator to obtain the membership vector S1 of qualitative data importance index and the membership vector S2 of quantitative data importance index. Finally, the fuzzy judgment set S is obtained by fuzzy transformation of the weight vector W10 formed by the weights of qualitative data importance index, quantitative data importance index and the weights of each dimension index under the impact index of data security risk (quantitative data importance index, qualitative data importance index and impact index of data security risk) and the evaluation matrix formed by S1, S2 and S3.

[0142] In some cases, the fuzzy synthesis operator used in the above fuzzy evaluation set is M(∧, ⊕) or M(., ⊕). Since these two operators take into account the role of comprehensive weight, strong comprehensive degree, high utilization rate of evaluation matrix and other factors during calculation, and adopt comprehensive normalization at the same time, they also reflect the idea of ​​weighted average.

[0143] ⊕ is a bounded sum operation, that is, an ordinary addition operation under bounded restrictions. ∧ means taking the smaller number, and . means multiplication.

[0144] The M(∧,⊕) operator acts on the weight vector W and the evaluation matrix R:

[0145]

[0146] The M(.,⊕) operator acts on the weight vector W and the evaluation matrix R:

[0147]

[0148] Among them, w i represents the i-th element corresponding to the weight vector W, r ikIt represents the element in the i-th row and k-th column of the evaluation matrix R. T represents the number of elements of W or the number of rows of R (the number of membership vectors, i.e., the number of indicators of a certain dimension in the indicator system), and Q represents the number of columns of R (the number of elements in a membership vector).

[0149] Step S16: Determine the level corresponding to the maximum element in the fuzzy evaluation set as the classification result of the target data.

[0150] In this embodiment, the level corresponding to the maximum element in the fuzzy evaluation set S calculated above is taken as the data classification result. The S vector is also the membership vector, and its specific form is as follows:

[0151] Level A Class B C-Class D-Class E-Class s1 s2 s3 s4 s5

[0152] S=(s1,s2,s3,s4,s5)

[0153] ms=max{s1,s2,s3,s4,s5}

[0154] Among them, ms represents the maximum value among {s1, s2, s3, s4, s5}, and the level corresponding to this value is extracted as the level of the target data, that is, based on the level extraction of the maximum membership, one of the levels A, B, C, D, and E can be obtained. If there are multiple elements equal to ms in S, the level closest to A is selected as the classification result of the target data.

[0155] The data classification indicator system of this embodiment can be applied to other indicator systems for data classification. Five levels are selected as the data level, but it is not limited to five levels, and can also cope with data classification tasks of other numbers of levels.

[0156] The method of this embodiment combines the fuzzy membership and election density weighting methods to perform data analysis, and uses fuzzy membership vectors to represent the evaluation results of leaf node indicators in the indicator system. It weights indicators at each level by iterative voting based on expert scoring to select important indicators and statistically analyze the indicator election density. It effectively avoids the inconsistency problem of some subjective weighting methods, simplifies and integrates the group decision-making mechanism, generates weight results through deterministic iterative optimization, and has low time and space complexity for system construction and algorithm operation, and is easy to accept, understand and operate. Furthermore, the experience of experts and evaluators is used to generate group decisions, integrate professionalism and group wisdom, optimize data definitions and compliance boundaries, quantify classification basis, and strengthen data classification management collaboration mechanisms. The method of this embodiment is applicable to industry-related competent (regulatory) departments, enterprises and institutions, and carries out data processing work such as data classification and grading, data processing services and data access authentication, providing an objective and effective technical solution.

[0157] Embodiment 2

[0158] Based on the above embodiment, this embodiment provides a data classification device, including:

[0159] An acquisition module, used to acquire target data to be classified;

[0160] The weighting module is used to assign weights to the indicators at each level in the preset hierarchical data classification indicator system for the target data. The data classification indicator system includes at least first-level indicators, second-level indicators and third-level indicators at different levels, wherein the first-level indicators are data classification indicators, the second-level indicators include qualitative data importance indicators, quantitative data importance indicators and data security risk impact indicators, and the third-level indicators include the lower-level indicators of the second-level indicators;

[0161] A first determining module is used to determine a data hierarchical set, where the data hierarchical set includes multiple levels;

[0162] A scoring module is used to score the underlying dimensional indicators in the data classification indicator system based on fuzzy membership, form corresponding underlying dimensional indicator membership vectors, and form an underlying dimensional indicator evaluation matrix based on the underlying dimensional indicator membership vectors;

[0163] A fuzzy transformation module is used to perform fuzzy transformation on the weight vector formed by the weights of the indicators of each dimension in the bottom layer and the evaluation matrix of the indicators of each dimension in the corresponding bottom layer through a fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the second bottom layer; in the second bottom layer, the evaluation matrix of the indicators of each dimension in each corresponding layer is formed by the membership vector of the indicators of each dimension in each same layer, and the weight vector formed by the weights of the indicators of each dimension in each corresponding layer and the evaluation matrix of the indicators of each dimension in each corresponding layer are fuzzy transformed through a fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, and the evaluation matrix of the indicators of each dimension in the previous layer of each corresponding layer is formed based on the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, iterate upward along the data classification indicator system to the second layer, and obtain the fuzzy judgment set through the fuzzy transformation of the weight vector formed by the weights of the indicators of each dimension in the second layer and the evaluation matrix of the indicators of each dimension in the second layer;

[0164] The second determination module is used to determine the level corresponding to the maximum element in the fuzzy evaluation set as the classification result of the target data.

[0165] In practical applications, the target data can be data from any industry field.

[0166] The data classification index system in this embodiment is divided into categories based on the importance of data in economic and social development, and the degree of harm caused to national security, economic operations, social order, public interests, organizational rights and interests, and personal rights and interests when the data is subject to security risks such as leakage, tampering, damage, or illegal acquisition, illegal use, and illegal sharing.

[0167] like Figure 2 As shown in the figure, the data classification indicator system includes indicators of different levels. Indicators of different levels constitute a hierarchical data classification indicator system. The first-level indicators are data classification indicators, and the lower levels are divided into qualitative data importance indicators, quantitative data importance indicators and data security risk impact indicators. At the same time, under these three types of indicators, the impact indicators of data classification are refined in multiple levels and dimensions, and data classification is carried out after identifying the above indicators.

[0168] Figure 2 In the data, the qualitative data importance index is divided into three types of indicators: business field, application entity and region. Among them, the business field is further divided into industry, business and content, the application entity is further divided into three levels of indicators such as people, groups, groups and systems, and the region is further divided into four levels of indicators such as places, geographical areas, administrative areas and socio-economic fields. The quantitative data importance index is divided into three levels of indicators such as accuracy, scale and coverage. The accuracy is further divided into four levels of indicators such as time accuracy and space accuracy. The scale is further divided into four levels of indicators such as storage volume, transaction volume and transmission volume. The coverage is further divided into four levels of indicators such as industry field, audience, geographical area and time. The impact of data security risks is further divided into three levels of indicators such as leakage, tampering, damage, illegal acquisition, illegal use and illegal sharing. The hierarchical data classification index system constructed based on these multi-level and multi-dimensional refined indicators can be applied to the data classification processing in this embodiment, so as to obtain scientific and accurate classification identification results, providing a basis for reasonable and effective data application and protection.

[0169] In some specific implementations, for target data, weights are assigned to indicators at each level in a preset hierarchical data grading indicator system, including: weights are assigned to indicators at each level in the data grading indicator system through iterative voting of important indicators by expert scoring and statistical election density. Further, it may include:

[0170] Determine the indicator sequence number set SN = {i|i = 1, ... m} of the current layer, determine the indicator set F = {fi|i∈SN} of the current layer, m represents the number of indicators in F, fi represents the i-th indicator in F;

[0171] Determine the expert group and the number of people n. In each round of voting, the experts who need to vote cannot abstain and must cast one vote; construct a valid voting indicator set E and initialize relevant information;

[0172] Enter the next iteration;

[0173] In the rth round, select the set of all experts ex that have the maximum voting density corresponding to the indicator in the r-1th round (r-1) Vote for the most important indicator in E. In the r-1th round, except for the expert set ex corresponding to the indicator with the maximum voting density in the rth round, (r-1) The votes of experts other than the above experts are still valid in round r, and these experts will not vote again in round r;

[0174] Count and store the number of votes for each indicator in the rth round E; calculate and store the voting correction coefficient for each indicator in the rth round E based on the number of elements in the rth round and r-1th round E;

[0175] Based on the maximum voting density of each indicator in E in the r-1th round, the voting correction coefficient of each indicator in E in the rth round, the number of votes obtained by the corresponding indicator in F in E in the rth round iteration, and the number of experts, the voting density of each indicator in E in the rth round is calculated and stored;

[0176] Get the set of numbers in F of the index with the highest voting density in E in round r Among them, mx r It represents the set of numbers in F of the index with the highest voting density in E in the rth round. The index with the highest voting density in E in the rth round corresponds to ec in F. r ={e rg |g∈mx r}, add the index corresponding to the maximum voting density and the elements in the tuple set formed by the voting density to the election density tuple set S, and the index corresponding to the maximum voting density and the tuple set formed by the voting density are recorded as et r ={(e rg ,vp rg )|g∈mx r};

[0177] Store the maximum voting density vpmax of each indicator in round r E r , Among them, mx r [0] indicates mx r The first element of the set, so vp r(mxr[0]) represents the maximum voting density in the rth round E;

[0178] Delete the index and its serial number corresponding to the maximum value of the voting density in round r: Delete ec from E r , E=E-ecr , remove mx from ESN r ,ESN=ESN-mx r ;

[0179] elnum (r+1) =||E||, the number of elements in E is saved to elnum (r+1) ;

[0180] Determine whether the number of elements in E is 1: If it is 1, calculate and store the voting density of the last indicator in E, and form a tuple set et of the maximum voting density indicator and density. (r+1) , et (r+1) =(e rlst ,vp rlst ), where lst represents the index number of the only index in E in F, et (r+1) In the rth round of calculation, the tuple set et (r+1) The unique indicator in is added to the election density tuple set S; otherwise, determine whether the element in E is 0:

[0181] If it is 0, the voting density vp in the tuples in the election density tuple set S is normalized to obtain the election density of each indicator, and the election density of each indicator is determined as the weight of each indicator; otherwise, enter the next round of iteration.

[0182] In some specific implementations, the calculation formula for the voting correction coefficient is as follows:

[0183]

[0184] In the formula, c r Indicates the correction coefficient for the rth round of voting, elnum r Indicates the number of elements in the rth round E, elnum (r-1) Represents the number of elements in E in the r-1th round.

[0185] The calculation formula for the voting density of each indicator in the rth round E is as follows:

[0186]

[0187] In the formula, vp rj represents the density of the jth index in F in E that is voted in the rth iteration, vpmax (r-1) It represents the maximum voting density of each indicator in the r-1th round E, and n represents the number of experts in the expert group.

[0188] In some specific implementations, the election density of each indicator is determined as the weight of each indicator, using the following calculation formula:

[0189]

[0190] In the formula, sp j represents the election density of the j-th indicator in F, S[j][2] represents the second element value of the j-th tuple in S, and S[x][2] represents the second element value of the x-th tuple in S.

[0191] It should be understood that the above only describes the weighting process of the same level of indicators. In this embodiment, each level of indicators in the data classification indicator system can be weighted by the above-mentioned expert scoring, iterative voting, and statistical voting density. No further details will be given here.

[0192] In this embodiment, the level is set to 5 levels, L = {level A, level B, level C, level D, level E}

[0193] The meaning of data importance indicators (including qualitative and quantitative indicators) is as follows:

[0194] Indicator level meaning Level A Very important Class B important C-Class More important D-Class generally E-Class unimportant

[0195] For the risk impact index, its meaning is:

[0196] Indicator level meaning Level A Suffering from data risks, the degree of harm is particularly serious Class B Suffering from data risks, the degree of harm is serious C-Class Suffering from data risks, the degree of harm is relatively large D-Class Suffering from data risk, the degree of harm is average E-Class Suffering from data risk, the degree of harm is small

[0197] In some specific implementations, the underlying dimensional indicators in the data classification indicator system are scored based on the fuzzy membership to form corresponding underlying dimensional indicator membership vectors, including:

[0198] Based on the importance of the target data in the underlying dimensional indicators under the qualitative data importance index, vote on the importance of the corresponding underlying dimensional indicators, count and quantify the voting results, and form the membership vector of the underlying dimensional indicators;

[0199] Based on the impact of the target data on the underlying dimensional indicators under the data security risk impact index, vote on the risk impact of the corresponding underlying dimensional indicators, count and quantify the voting results, and form the membership vector of the underlying dimensional indicators; and

[0200] Get the values ​​of the underlying dimensional indicators under the quantitative data importance index, use the preset membership function corresponding to each level to calculate the output value of the corresponding level and normalize the output value to form the membership vector of the underlying dimensional indicators.

[0201] The fuzzy transformation process of the weights of each layer of indicators and the evaluation matrix is ​​as follows:

[0202]

[0203] Among them, S1 represents the membership vector of the qualitative data importance index, W8 represents the weight vector formed by the weights of each dimensional index (business field, application entity, region) under the qualitative data importance index, R1 represents the evaluation matrix formed by the membership vector of the underlying dimensional indicators (industry, business, content) under the business field, W1 represents the weight vector formed by the weights of the underlying dimensional indicators under the business field, R2 represents the evaluation matrix formed by the membership vector of the underlying dimensional indicators (people, groups, groups, systems) under the application entity, W2 represents the weight vector formed by the weights of the underlying dimensional indicators under the application entity, R3 represents the evaluation matrix formed by the membership vector of the underlying dimensional indicators under the region, W3 represents the weight vector formed by the weights of the underlying dimensional indicators under the region (place, geographical area, administrative area, socio-economic field), is the fuzzy synthesis operator.

[0204]

[0205] Among them, S2 represents the membership vector of the quantitative data importance index, W9 represents the weight vector formed by the weights of each dimensional index (precision, scale, coverage) under the quantitative data importance index, R4 represents the evaluation matrix formed by the membership vectors of the underlying dimensional indicators (time accuracy, spatial accuracy) under accuracy, W4 represents the weight matrix formed by the weights of the underlying dimensional indicators under accuracy, R5 represents the evaluation matrix formed by the membership vectors of the underlying dimensional indicators (storage volume, transaction volume, transmission volume) under scale, W5 represents the weight vector formed by the weights of the underlying dimensional indicators under scale, R6 represents the evaluation matrix formed by the membership vectors of the underlying dimensional indicators (industry field, audience, geographical area, time) under coverage, and W6 represents the weight vector formed by the weights of the underlying dimensional indicators under coverage.

[0206]

[0207] Among them, S3 represents the membership vector of the data security risk impact index, W7 represents the weight vector formed by the weights of the underlying dimensional indicators (leakage, tampering, damage, illegal acquisition, illegal use, illegal sharing) under the data security risk impact index, and R7 represents the evaluation matrix formed by the membership vectors of the underlying dimensional indicators under the data security risk impact index.

[0208]

[0209] Among them, S represents the fuzzy judgment set of data classification indicators, and W10 represents the weight vector formed by the weights of each dimensional indicator under the data classification indicator (quantitative data importance indicator, qualitative data importance indicator and data security risk impact indicator).

[0210] In some cases, the fuzzy synthesis operator used in the above fuzzy evaluation set is M(∧, ⊕) or M(., ⊕). Since these two operators take into account the role of comprehensive weight, strong comprehensive degree, high utilization rate of evaluation matrix and other factors during calculation, and adopt comprehensive normalization at the same time, they also reflect the idea of ​​weighted average.

[0211] ⊕ is a bounded sum operation, that is, an ordinary addition operation under bounded restrictions. ∧ means taking the smaller number, and . means multiplication.

[0212] The M(∧,⊕) operator acts on the weight vector W and the evaluation matrix R:

[0213]

[0214] The M(.,⊕) operator acts on the weight vector W and the evaluation matrix R:

[0215]

[0216] Among them, w i represents the i-th element corresponding to the weight vector W, r ik It represents the element in the i-th row and k-th column of the evaluation matrix R. T represents the number of elements of W or the number of rows of R (the number of membership vectors, i.e., the number of indicators of a certain dimension in the indicator system), and Q represents the number of columns of R (the number of elements in a membership vector).

[0217] It should be understood that the device of this embodiment has all the beneficial effects of the first embodiment.

[0218] Embodiment 3

[0219] On the basis of the above embodiments, this embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory, and the processor executes the computer program to implement the steps of the method described in the above embodiments.

[0220] In some implementations of this embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in the above embodiment are implemented.

[0221] In some implementations of this embodiment, a computer program product is provided, including a computer program / instructions, and when the computer program is executed by a processor, the steps of the method described in the above embodiment are implemented.

[0222] The processor may include, but is not limited to, one or more processors or microprocessors. Each processor may be an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a controller, a microcontroller, a microprocessor or other electronic components to execute the method in the above embodiments.

[0223] The computer-readable storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, and the computer-readable storage medium may include but is not limited to, for example, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, computer storage media (e.g., hard disk, floppy disk, solid-state drive, removable disk, CD-ROM, DVD-ROM, Blu-ray disc, etc.).

[0224] The computer-readable storage medium may also store at least one computer executable program / instruction, which may be, for example, a computer-readable instruction. The computer-readable storage medium includes, but is not limited to, for example, a volatile memory and / or a non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The computer-readable storage medium may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc. For example, a non-transitory computer-readable storage medium may be connected to a computing device such as a computer, and then, when the computing device runs the computer-readable instructions stored on the computer-readable storage medium, the various methods described above may be performed.

[0225] In addition, the computer device may also include (but not limited to) a data bus, an input / output (I / O) bus, a display, and input / output devices (eg, keyboard, mouse, speaker, etc.), etc.

[0226] The processor may communicate with external devices via an I / O bus via a wired or wireless network.

[0227] In one embodiment, the at least one computer executable instruction may also be compiled into or constitute a software product / computer program product, wherein one or more computer executable instructions are executed by a processor to perform the various functions and / or method steps in the embodiments described in the present technology.

[0228] Embodiment 4

[0229] This embodiment provides an application example of the above embodiment. Figure 7 shown.

[0230] The data classification method in the aforementioned embodiment is implemented as a software system.

[0231] Submit and display the following information on the software front end:

[0232] Expert Collectionex (r-1) In the rth round of voting, the experts submit their votes for the most important indicators in E according to their experience, and apply the election density weighting method to calculate the weights of indicators in each dimension and level. (r-1) The votes of experts other than for the indicators are still valid in round r, and these experts cannot vote again in round r;

[0233] The preset membership functions of each level of the objective indicator are set, and the administrator fills in the objective indicator input;

[0234] Each evaluator submits an evaluation vote for the subjective indicators.

[0235] The backend performs the following information processing:

[0236] Assign weights to each indicator, including:

[0237] Count and store the number of votes obtained by the jth indicator of F in the rth round E in the voting;

[0238] Calculate and store the voting correction coefficients of each indicator in the rth round E;

[0239] Calculate and store the voting density of each indicator in round r of E;

[0240] Get the set of serial numbers in F of the index with the largest voting density in E in round r;

[0241] Update E, ESN, vpmax, elnum, S, etc.;

[0242] Determine whether the number of elements in E is 1, 0, or greater than 1:

[0243] If it is 1, the voting density of the last indicator in E is calculated and stored; the voting density vp in the tuple in S is normalized to obtain the election density of each indicator, which is the weight of each indicator;

[0244] If is 0, the voting density vp in the tuple in S is directly normalized to obtain the election density of each indicator, which is the weight of each indicator;

[0245] If it is greater than 1, feedback will be given to the experts to continue voting;

[0246] For subjective indicators, the voting results are counted and quantified, and the total is normalized to form a membership vector;

[0247] For the objective indicator system, the indicator value is input into the membership function of each level, the membership function value of each level is generated, and the total is normalized to form the membership function;

[0248] Calculate the fuzzy evaluation set S (membership vector) and determine the data level based on the maximum membership principle.

[0249] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products, or devices.

[0250] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0251] In the embodiments provided in the present disclosure, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the above-mentioned module, program segment or a part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0252] It should be noted that in the present disclosure, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element limited by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0253] Although the embodiments disclosed in the present disclosure are as above, the above contents are only embodiments adopted for facilitating the understanding of the present disclosure and are not intended to limit the present disclosure. Any technician in the technical field to which the present disclosure belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present disclosure, but the scope of patent protection of the present disclosure shall still be subject to the scope defined in the attached claims.

Claims

1. A data classification method, characterized in that: include: Obtain target data to be classified; For the target data, weights are assigned to indicators at each level in a preset hierarchical data classification indicator system, wherein the data classification indicator system includes at least first-level indicators, second-level indicators and third-level indicators at different levels, wherein the first-level indicators are data classification indicators, the second-level indicators include qualitative data importance indicators, quantitative data importance indicators and data security risk impact indicators, and the third-level indicators include lower-level indicators of the second-level indicators; determining a data hierarchical set, the data hierarchical set comprising a plurality of levels; Scoring the underlying dimensional indicators in the data classification indicator system based on fuzzy membership to form corresponding underlying dimensional indicator membership vectors, and forming an underlying dimensional indicator evaluation matrix based on the underlying dimensional indicator membership vectors; The weight vector formed by the weights of the indicators of each dimension in the bottom layer is fuzzily transformed with the evaluation matrix of the indicators of each dimension in the corresponding bottom layer by using the fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the second bottom layer; in the second bottom layer, the evaluation matrix of the indicators of each dimension in each corresponding layer is formed by the membership vector of the indicators of each dimension in each same layer, and the weight vector formed by the weights of the indicators of each dimension in each corresponding layer is fuzzily transformed with the evaluation matrix of the indicators of each dimension in each corresponding layer by using the fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, and the evaluation matrix of the indicators of each dimension in the previous layer of each corresponding layer is formed based on the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, iterate upward along the data classification indicator system to the second layer, and obtain the fuzzy judgment set by fuzzy transformation of the weight vector formed by the weights of the indicators of each dimension in the second layer and the evaluation matrix of the indicators of each dimension in the second layer; The level corresponding to the maximum element in the fuzzy evaluation set is determined as the classification result of the target data.

2. The data classification method according to claim 1, characterized in that: For the target data, weights are assigned to each level of indicators in the preset hierarchical data classification indicator system, including: Through iterative voting of important indicators by expert scoring and statistical election density, weights are assigned to indicators at each level in the data grading indicator system.

3. The data classification method according to claim 2, characterized in that: Based on the expert scoring, iterative voting for important indicators and statistical voting density, weights are assigned to each level of indicators in the preset hierarchical data classification indicator system, including: Determine the index sequence number set SN of the current layer and determine the index set F of the current layer; Determine the expert group and its number. Experts who are required to vote in each round of voting may not abstain and must cast one vote; Construct a valid voting indicator set E and initialize related parameters to enter the next round of iteration; In the rth round, the set of all experts who voted for the index with the maximum voting density in the r-1th round is selected to vote on the most important index in E. In the r-1th round, the votes of experts other than the set of experts who voted for the index with the maximum voting density in the rth round are still valid in the rth round, and these experts will not vote again in the rth round; Count and store the votes for each indicator in round r of E; Based on the number of elements in E in the rth round and r-1th round, the voting correction coefficient of each indicator in E in the rth round is calculated and stored; Based on the maximum voting density of each indicator in E in the r-1th round, the voting correction coefficient of each indicator in E in the rth round, the number of votes obtained by the corresponding indicator in F in E in the rth round iteration, and the number of experts, the voting density of each indicator in E in the rth round is calculated and stored; Get the set of serial numbers of the index with the largest voting density in E in the rth round in F, correspond the index with the largest voting density in E in the rth round to F, add the index corresponding to the maximum voting density and the elements in the tuple set formed by the voting density to the election density tuple set S; Store the maximum value of the voting density of each indicator in the rth round E, and delete the indicator and its serial number corresponding to the maximum value of the voting density in the rth round in E; Determine whether the number of elements in E is 1: If it is 1, calculate and store the voting density of the last indicator in E, and form a tuple set of the maximum voting density indicator and density, and add the only indicator in the tuple set to the election density tuple set S; Otherwise, determine whether the element in E is 0: If it is 0, the voting density vp in the tuples in the election density tuple set S is normalized to obtain the election density of each indicator, and the election density of each indicator is determined as the weight of each indicator; Otherwise, proceed to the next iteration.

4. The data classification method according to claim 3, characterized in that: The calculation formula of the voting correction coefficient is as follows: In the formula, c r Indicates the correction coefficient for the rth round of voting, elnum r Indicates the number of elements in the rth round E, elnum (r-1) Represents the number of elements in E in the r-1th round; The calculation formula for the voting density of each indicator in the rth round E is as follows: In the formula, vp rj represents the density of the jth index in F in E that is voted in the rth iteration, vpmax (r-1) represents the maximum value of the voting density of each indicator in the r-1th round E, and n represents the number of experts in the expert group; The election density of each indicator is determined as the weight of each indicator, using the following calculation formula: In the formula, sp j represents the election density of the j-th indicator in F, S[j][2] represents the second element value of the j-th tuple in S, and S[x][2] represents the second element value of the x-th tuple in S.

5. The data classification method according to claim 1, characterized in that: Based on the fuzzy membership, the underlying dimensional indicators in the data classification indicator system are scored to form corresponding underlying dimensional indicator membership vectors, including: Based on the importance of the target data in the underlying dimensional indicators under the qualitative data importance index, vote on the importance of the corresponding underlying dimensional indicators, count and quantify the voting results, and form the corresponding underlying dimensional indicator membership vector; Based on the impact degree of the target data on the underlying dimensional indicators under the data security risk impact degree indicator, vote on the risk impact degree of the corresponding underlying dimensional indicators, count and quantify the voting results, and form the corresponding underlying dimensional indicator membership vector; Get the values ​​of the underlying dimensional indicators under the quantitative data importance index, use the preset membership function corresponding to each level to calculate the output value of the corresponding level and normalize the output value to form the corresponding membership vector of the underlying dimensional indicators.

6. The data classification method according to claim 5, characterized in that: The preset membership function includes one or more of an S-type membership function, a double S-type membership function, a Z-type membership function, a trapezoidal membership function, a triangle membership function, a Gaussian membership function, a Gaussian combination membership function, and a generalized bell-shaped membership function.

7. A data classification device, characterized in that: include: An acquisition module, used to acquire target data to be classified; A weighting module, for assigning weights to indicators at each level in a preset hierarchical data classification indicator system for the target data, wherein the data classification indicator system includes at least primary indicators, secondary indicators and tertiary indicators at different levels, wherein the primary indicators are data classification indicators, the secondary indicators include qualitative data importance indicators, quantitative data importance indicators and data security risk impact indicators, and the tertiary indicators include lower-level indicators of the secondary indicators; A first determining module, configured to determine a data hierarchical set, wherein the data hierarchical set includes a plurality of levels; A scoring module is used to score the underlying dimensional indicators in the data classification indicator system based on fuzzy membership, form corresponding underlying dimensional indicator membership vectors, and form an underlying dimensional indicator evaluation matrix based on the underlying dimensional indicator membership vectors; A fuzzy transformation module is used to perform fuzzy transformation on the weight vector formed by the weights of the indicators of each dimension in the bottom layer and the evaluation matrix of the indicators of each dimension in the corresponding bottom layer through a fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the second bottom layer; in the second bottom layer, the evaluation matrix of the indicators of each dimension in each corresponding layer is formed by the membership vector of the indicators of each dimension in each same layer, and the weight vector formed by the weights of the indicators of each dimension in each corresponding layer and the evaluation matrix of the indicators of each dimension in each corresponding layer are fuzzy transformed through a fuzzy synthesis operator to obtain the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, and the evaluation matrix of the indicators of each dimension in the previous layer of each corresponding layer is formed based on the membership vector of the indicators of each dimension in the previous layer of each corresponding layer, iterate upward along the data classification indicator system to the second layer, and obtain the fuzzy judgment set through the fuzzy transformation of the weight vector formed by the weights of the indicators of each dimension in the second layer and the evaluation matrix of the indicators of each dimension in the second layer; The second determination module is used to determine the level corresponding to the maximum element in the fuzzy evaluation set as the classification result of the target data.

8. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images